Toward a Systematic Understanding of Photodetectors Based on

Aug 22, 2008 - Toward a Systematic Understanding of Photodetectors Based on Individual Metal Oxide Nanowires. Joan Daniel Prades*, Roman Jimenez-Diaz,...
0 downloads 6 Views 926KB Size
J. Phys. Chem. C 2008, 112, 14639–14644

14639

Toward a Systematic Understanding of Photodetectors Based on Individual Metal Oxide Nanowires Joan Daniel Prades,*,† Roman Jimenez-Diaz,† Francisco Hernandez-Ramirez, Luis Fernandez-Romero,† Teresa Andreu,† Albert Cirera,† Albert Romano-Rodriguez,† Albert Cornet,† Joan Ramon Morante,† Sven Barth,‡ and Sanjay Mathur*,‡,§ EME/XaRMAE/IN2UB, Departament d’Electronica, UniVersitat de Barcelona, C/ Marti i Franques 1, Barcelona, E-08028, Spain, Nanocrystalline Materials and Thin Film Systems, Leibniz-Institute of New Materials, Saarbruecken, D-66123, Germany, and Department of Inorganic Chemistry, UniVersity of Cologne, Cologne, D-50939, Germany ReceiVed: May 20, 2008; ReVised Manuscript ReceiVed: July 3, 2008

We present a set of criteria to optimize photodetectors based on n-type metal oxide nanowires and a comparison methodology capable of overcoming the present lack of systematic studies dealing with such devices. The response of photoconductors is enhanced following different fabrication strategies, such as diminishing the distance between the electrical contacts, increasing the width of the photoactive area, or improving the electrical mobility of the nanomaterials. The validity of the theoretical background is verified by experimental results obtained with devices based on ZnO nanowires. The performances of our devices show that the normalized gain of single ZnO nanowire-based photodetectors exceeds those of thin films. Introduction Metal oxide nanowires are gaining growing interest as photodetectors due to their potential applications in gas sensing and optoelectronics.1 Although preliminary works revealed promising results,1-11 further research is necessary in order to reach complete control of their photosensing properties. Among all n-type metal oxide nanomaterials, photoresponses of ZnO and SnO2 nanowires have been widely studied;1-11 however, the lack of well-established fabrication methodologies and standardized procedures obfuscates a comparison of experimental results. For instance, photoconductive gains (Gph) ranging from 102 to 108 4,9,10 and response times (τ) between milliseconds and hours were reported for ZnO nanowires,1,2,4,7,10 which possibly result from different experimental conditions and device geometry used in all of these studies. Here, we present some systematic strategies for enhancing the response of photodetectors based on nanowires and a methodology for the facile comparison of the measured photoresponses. These strategies are fundamental to the potential of this technology in real devices and applications. Experimental Methods ZnO nanowires were fabricated via a vapor-phase carbothermal transport process inside an Atomate’s chemical vapor deposition (CVD) system. The source material was a 1:1 molar mixture of commercial ZnO (metal basis, 99.99%) and graphite powder (crystalline, 300 mesh, 99%) from Alfa Aesar. Gold nanoparticles were used as catalytic islands on thermally oxidized Si wafers.12 Uniform nanowires were obtained with mean radius ) 90 ( 15 nm and lengths up to 30 µm. Some of them were electrically contacted to platinum microelectrodes using a FEI Dual-Beam Strata 235 FIB instrument * Corresponding authors: [email protected], [email protected]. † Universitat de Barcelona. ‡ Leibniz-Institute of New Materials. § University of Cologne.

Figure 1. Diagram of a metal-oxide structure with an arbitrary volume of length L, width W, and thickness T, under a photon flux Φph. Photocarrier generation is induced in the upper layer of the structure until a depth R-1 and leads to photoconduction (jph) under polarization.

following a nanolithography process explained elsewhere.13-15 Electrical measurements were performed with the help of a selfmade electronic circuit designed to ensure low current levels and to avoid undesired fluctuations16 inside a homemade chamber. Photoresponse was excited using both UV LEDs and a UV lamp. UV LEDs were centered on λ ) 340 ( 10 nm and λ ) 385 ( 15 nm (Seoul Optodevices T9F34C and Purple-Hi E1L5M-4P0A2), whereas the UV lamp was a HAMAMATSU LC8 light source with a type[-01] Hg-Xe lamp enhanced for the line at λ ) 365 nm. Light intensity impinging on the nanowires was determined with the help of a thermopile detector

10.1021/jp804614q CCC: $40.75  2008 American Chemical Society Published on Web 08/22/2008

14640 J. Phys. Chem. C, Vol. 112, No. 37, 2008

Prades et al.

(FieldMax-TOP). Up to 10 devices based on single ZnO nanowires were fabricated, and their photoresponses were studied as a function of different experimental parameters to validate the theoretical discussion. Two-probe I-V measurements revealed symmetric responses in great accordance with results published elsewhere.17 The effective voltage drop along the nanowires was determined by decoupling the rectifying contact from the purely resistive contribution of the nanowires following the procedure described elsewhere.17 To evaluate the role of surface contribution on the photoresesponse of these devices, some samples were coated with a 475 ( 50 nm thick layer of PMMA (2 wt % poly (methyl methacrylate) in dichloromethane) by spin coating. All experiments were performed in real air atmosphere. Results and Discussion 1. Theoretical Background. Photodetectors based on single n-type metal oxide nanowires, modeled as an arbitrary volume of length L, width W, and thickness T (Figure 1), can be studied using the fundamental principles ruling light carrier generation on semiconductors.18 Thus, current density jph is given by the following equation,

jph ) q∆nphV

(1)

where q is the elemental charge, ∆nph is the concentration of generated carriers, and V is their velocity. ∆nph can be also written as shown in eq 2,

∆nph )

ηF ηF τ τ≈ Vph R-1WL

(2)

where η is the quantum efficiency of carrier generation by one photon, F is the absorption rate of photons, τ is the carrier lifetime, and Vph is the photogeneration volume (Figure 1). The approximation in eq 2 assumes a constant carrier generation profile until the depth R-1, where R is the absorption coefficient of the metal oxide at one fixed wavelength (see the Supporting Information). ∆nph and τ are also related through the following continuity equation,18

∆nph ∂∆nph ) gph ∂t τ

(3)

where gph is the generation rate of charge carriers under illumination. Assuming that the concentration of electrons ∆nph is independent of τ, the photocurrent dynamics at the rising (time constant τr) and falling (time constant τd) edges are given by (see Supporting Information),

iph(t) ) Iph(1 - e iph(t) ) Iphe

-t⁄

τr

-t⁄ τd

)

(4) (5)

where Iph is the photocurrent at the steady-state. However, if an external field E is applied parallel to the longitudinal axis of ohmic nanowires, then the velocity of carriers (V) can be expressed in terms of the applied voltage (V) using the following formula,

V)µ*E)

µ*V L

(6)

where µ* is the effective carrier mobility. According to the Matthiessen’s rule, µ* can be divided into the bulk (µB) and surface (µS) contribution according to eq 7.18

1 1 1 + ) µ* µB µS

(7)

The absorption rate of photons F when metal oxide nanowires are exposed to a flux of Φph photons can be also expressed by eq 8,

F ) βΦphWL

(8)

where β is the fraction of photons not reflected by the surface, and WL is the effective area of one nanowire (see Figure 1). Therefore, using eqs 2, 6, and 8, current density jph can be rewritten as eq 9

jph ) q

1 βητµ * VΦph LR-1

(9)

To evaluate the total photogenerated current Iph, which is the experimental response of real devices, we assume that nanowires are thick enough to absorb all the incident photons. That is to say,

T g R-1

(10)

Therefore, it can be deduced that thinner nanowires (T < R-1) will lead to lower photoresponses. On the contrary, the use of thicker nanowires (T . R-1) will not imply a further signal enhancement. For instance, the penetration depth R-1 of near-UV photons (wavelength from 400 to 250 nm) in ZnO is almost constant at 50 nm.19 Thus, ZnO nanowires with radii slightly above r ≈ 25 nm should be used to maximize photoresponse to UV-light in this wavelength range. If the constant absorption profile approximation is maintained, photocurrent Iph in nanowires that satisfy eq10 is given by eq 11,

W Iph ) jph(R-1W) ) q βητµ * VΦph L

(11)

where three different contributions are clearly identified. The first one is related to geometric parameters of the device (W/ L), the second one to the intrinsic properties of nanowires (βητµ*) and the third one only depends on the experimental conditions (VΦph). The performance of these devices can be also analyzed in terms of their photoconductive gain Gph, which is defined in eq 12,18

Gph ≡

Iph 1 ≈ ητµV qF L2

(12)

where identical types of contributions are involved. 2. Experimental Validation and Discussion. 2.1. Geometrical Aspects. Concerning the geometry of photodetectors, eq 11 implies enhanced Iph with increasing width (W) of the photoactive area. To verify this, the responses of single ZnO nanowires and lamellae were compared.20 ZnO lamellae were randomly obtained during the synthesis process of ZnO nanowires when the density of catalytic gold nanoparticles is very high (Figure 2). Other authors have postulated that the origin of these structures could be related to variations of the oxygen concentration during synthesis.20 ZnO lamellae presented thicknesses (T) comparable to the diameter of the nanowires (2r ≈ 220 nm), but their widths (W) were up to 10 times larger, allowing us to confirm the enhancement of Iph with increasing photoactive areas (Figure 2a). To compare the UV photosensing performances of individual ZnO nanowires with ZnO lamellae, identical conditions were used in all experiments (Φph ) 3.3 × 1018 ph m-2

Photodetectors Based on Individual Metal Oxide Nanowires

Figure 2. (a) ZnO lamella of length LL ) 3.9 µm and average width around WL ≈ 1000 nm contacted with FIB nanolithography techniques. (Inset) SEM image of the mixture of ZnO nanowires and ZnO lamellae obtained after the synthesis process. (b) Normalized photoresponse I˜ph to the distance between electrical contacts L of one ZnO lamella and one ZnO nanowire (LNW ) 13 µm) under identical conditions (Φph ) 3.3 × 1018 ph m-2 s-1; λ ) 340 ( 10 nm). Iph is higher with the lamella due to a larger value of W.

s-1; λ ) 340 ( 10 nm) and a comparable photocurrent (I˜ph) value is obtained by normalization to the interelectrode distances (L). ˜

Iph )

Iph 1

⁄L

(13)

The photoresponse [Figure 2] obtained with a nanowire was 4.7 times lower than the one measured with the lamella shown in Figure 2a. The average width of this lamellar structure was WL ≈ 1000 nm, whereas the nanowire showed WL ≈ 220 nm in diameter (WL/WNW ≈ 4.5). In accordance to eq 11, the higher active area of ZnO lamellae caused enhanced photocurrent values I˜ph as predicted by geometrical factors and shown by the presented values (WL/WNW ≈ I˜ph(L)I˜ph(NW)). Another convenient way to increase the width of the photoactive area is using multi-nanowire-based configurations. These devices can be realized by self-assembly techniques, such as dielectrophoresis, to electrically contact large amounts of nanowires in parallel.21,22 This fabrication methodology prevents parasitic effects arising from uncontrolled nanowire-nanowire contacts, and the resulting devices admit higher currents (with simpler conditioning electronics) without damaging the nanowires16 (see the Supporting Information). It is noteworthy that, according to eq 12, the photoconductive gain Gph obtained with these multi-nanowire configurations is equivalent to the gain provided by one single nanowire, if all of them are identical. The distance between contacts (L) also determines the response of the photodetector since both Iph and Gph increase inversely with this parameter. L not only influences the photocapture area (WL) but also determines the effective electric

J. Phys. Chem. C, Vol. 112, No. 37, 2008 14641

Figure 3. (a) ZnO nanowire with radius r ) 90 ( 10 nm contacted to five contacts fabricated with FIB nanolithography techniques. (b) Photoresponse Iph and photoconductive gain Gph (inset) of the ZnO nanowire as a function of the distance L between electrical contacts obtained with Φph ) 3.3 × 1018 ph m-2s-1, λ ) 340 ( 10 nm, and V ) 1 V.

field (E) inside the nanowire due to the bias voltage (V) applied externally. Indeed, this second aspect dominates the overall contribution of L to the photoresponse (see eqs 11 and 12). The dependence of Iph and Gph on L was experimentally confirmed by fabricating five electrical contacts separated at different distances on an individual ZnO nanowire (Figure 3a), and measuring the UV photoresponse between different pairs of electrodes. It was experimentally found that Iph and Gph increase inversely with the distance between contacts L, in accordance with eqs 11 and 12 (Figure 3b). Therefore, it can be concluded that higher-gain photodetectors are obtained by diminishing this fabrication-related parameter. The lower limit for L will strictly depend on the precision of the nanolithography technique and other size-associated phenomena such as diffraction, if L approaches the wavelength of photons, or uncontrolled degradation effects produced when the rupture electrical field of the metal oxide is overcome (see the Supporting Information). To exemplify the later point, it can be roughly estimated that nanowires contacted between two electrodes with a separation of only 50 nm23 and polarized at 5 V 1,10 will be subjected to electrical fields as high as 1 MV/cm. 2.2. Intrinsic Properties of the Semiconductor. In addition to geometrical factors, the dependence of Iph on intrinsic material properties, such as η , τ, µ* and β , have to be considered (see eq 11). The spectral response of photodetectors is determined by the quantum efficiency η, which was observed to increase by up to 3 orders of magnitude when photons with energies above the bandgap interact with these devices, compared to typical responses obtained with sub-bandgap photons.24 The variation of quantum efficiency η as a function of incident light wavelength (λ) was investigated. In Figure 7, it is demonstrated how Iph scaled up when above-bandgap photons (λ ) 340 ( 10 nm) interacted with a single ZnO nanowire-based device, compared to low Iph response obtained with sub-bandgap photons (λ ) 385 ( 15 nm). This result shows the importance of tuning the bandgap of such photodetectors to select their

14642 J. Phys. Chem. C, Vol. 112, No. 37, 2008

Figure 4. Dynamic behavior of the photoresponse Iph measured with an individual ZnO nanowire when a UV pulse is applied (dashed line) (Φph ) 3.3 × 1018 ph m-2 s-1; λ ) 340 ( 10 nm; V ) 1 V). The response fits with theoretically deduced exponential laws and exhibits time constants of τr ) 170 ( 3 s and τd ) 300 ( 8 s.

active/blind spectral regions.11,18,24 It is noteworthy that the bandgap edge of nanowires depends not only on the material but also on their dimensions.11,18 Thus, controlling the radii of nanowires is critical to tune the spectral sensitivity of the final devices. The photogenerated carrier lifetime (τ) is the second parameter directly related to the intrinsic properties of nanowires, which is known to be higher in nanomaterials compared to bulk due to the large surface-to-volume ratio and the formation of deep level surface states.18,25 For metal oxide nanowires, it is generally accepted that photocarrier relaxation dynamics consists of an initial decay process in the nanosecond range, explained by the fast carrier thermalization and hole-trapping by surface states, followed by a slow decay dependent on the surrounding atmosphere and the nanowire surface coating.3,4,26,27 This second process, with characteristic time constants in seconds, dominates the final response of nanowire-based photodetectors. For this reason, the carrier lifetime contribution τ to the photoresponse Iph (see eq 11) can be modified by controlling the surface interactions of this type of nanowires. Time-resolved measurements allowed us to estimate the rising time constant τr and falling time constant τdexhibited by our devices (Figure 4). The evolution in time (iph(t)) was in accordance with the theoretical model (eqs 4 and 5), whereby values close to a few minutes were found. The third parameter related to the intrinsic properties of nanowires is the electrical mobility µ/, which is known to be dependent on their radii. In the case of ZnO, mobility values ranging from 2 to 30 cm2/(V s) were reported for nanowires with radii below r ≈ 100 nm.28-31 The diminished mobility increases up to the bulk value (∼200 cm2/(V s)) in thicker nanowires.10 This behavior is attributed to scattering and trapping of the electrons by surface defect states and becomes more significant in thin nanowires possessing higher surfaceto-volume ratios. Thus, thicker nanowires are convenient to obtain optimal devices for µ* optimization due to minimized surface contributions µS (eq 7). The limitation introduced by the dependence of µ* with radius can be also circumvented by passivating the nanowire surface, which is reported to dramatically increase the mobility of ZnO nanowires (up to 1.000 cm2/ (V s)).31-34 It was demonstrated that the mobility of ZnO nanowires dramatically increased (up to ∼1.000 cm2/(V s)) with Si3N4/ SiO2,31 polyimide,32 poly(methyl methacrylate) (PMMA),33 and polyacrylonitrile34 coatings. Therefore, we covered some devices

Prades et al.

Figure 5. Photoresponse (I˜ph ≡ IL) normalized to the distance between electrical contacts L of one single ZnO nanowire (LNW ) 13 µm) before and after coating with PMMA under identical illumination (Φph ) 3.3 × 1018 ph m-2 s-1; λ ) 340 ( 10 nm). I˜ph is higher after coating due to surface passivation induced higher µ/.

Figure 6. Linear dependency of photoresponses Iph on the applied voltage V of an individual ZnO nanowire (under two different photon fluxes Φph).

Figure 7. Photoresponse Iph of an individual ZnO nanowire under different photon fluxes Φph at constant polarization voltage (V ) 1 V). A linear dependence between these two parameters is experimentally observed. Remarkably, photoresponse Iph is highly dependent on the energy of the incident photons.

with PMMA by spin coating after eliminating the surface contamination with oxygen plasma. The passivation layers were approximately 500 nm thick, and presented high UV transparency (optical transmittance above 92% from 300 to 800 nm) and extremely low conductance (below 0.1 nS) independent of the illumination. The photoresponse of these devices was ∼9.5 times higher after coating (see Figure 5). To elucidate this point, the mobility of the nanowires15 before and after passivation were estitmated to be µ* ≈ 3-5 cm2/(V s) and µ*(PMMA) ≈ 40-53 cm2/(V s), respectively. This improvement in the nanowire mobility (factor of ∼10) and the high transmittance of the PMMA layer explain the increased photoresponse according to eq 11.

Photodetectors Based on Individual Metal Oxide Nanowires The last intrinsic parameter of nanowires to be considered in this work is the fraction of photons (β) not reflected by the surface of the metal oxide, which was recently demonstrated to be lower in photodetectors based on aligned nanowires compared to thin films.35 2.3. Working Conditions. As far as the experimental conditions are concerned, it can be expected from eqs 11 and 12 that photoresponse raises linearly with applied voltage V and flux of photons Φph. Figures 6 and 7 demonstrate the linear increase in device response in dependence to the two parameters and UV-light according to theory. This feature complicates comparing most of the reported results, since different experimental conditions were used in these experiments.1-10 For this reason we propose a more generic way to express eq 12,

gph ≡

L2Gph ≈ ητµ* V

(14)

which is a normalized photoconductive gain, independent of the device geometry and the experimental conditions. Apart from the photoconductive gain, the dynamic behavior of these devices is also important for optimizing real devices. Their low-pass bandwidth can be defined as BW ≈ (2πτ)-1. Thus, the normalized gain per bandwidth becomes

gphBW ≈

ηµ* 2π

(15)

From equation 15, it can be observed that the quantum efficiency η and mobility µ/ are the key parameters to evaluate the overall performance of these photodetectors. 3. Comparison to State-of-the-art Devices. To substantiate and validate the performances of our nanowire-based devices, the photoresponse values were critically compared with previously reported results. Soci et al.10 recently reported photoconductive gain values of Gph,lit ) 5 × 107 in devices with Llit ) 1 µm and lit ) 110 nm polarized at Vlit ) 5 V for equivalent illumination conditions, which is much higher than the typical Gph values measured in our experiments. For example, Gph ) 6 × 103 was obtained with V ) 1 V, L ≈ 13 µm, and r ≈ 110 nm. However, an accurate comparison of these two devices can be only achieved if the photoconductive gain Gph is rewritten as shown in equation 14. If gph is calculated, we find that gph,lit ) 10-5 m2/V and gph ) 10-6 m2/V, showing that the literature value is only 1 order of magnitude higher than ours. This discrepancy can be explained considering the mobility µ/ and lifetime τd of nanowires. On one hand, our ZnO nanowires have response time constants close to τ ≈ 300s (Figure 4) and exhibited mobility values of ∼3 cm2/(V s), which was estimated following a procedure described elsewhere.15 On the other hand, * Soci and co-workers reported τlit ≈ 33 s and µlit ≈ 270 cm2/(V * s), whose product (τlitµlit) is a factor 10 time higher than ours (τµ*), which justifies the divergence of values in gph (see eq 14). Fully consistent results were obtained with the rest of devices under test, demonstrating that gain normalization of photodetectors to the geometry and polarization conditions is necessary for adequate comparison. Moreover, comparison of gph values allows the determination of which intrinsic properties of the photoactive material must be improved to enhance the performance of future devices. For example, according to the experimental results, the electron mobility µ/ should be improved in our case. From all of these results, it can be concluded that the experimental responses obtained with photodetectors based on single nanowires are precisely modeled by the heresummarized theoretical approach.

J. Phys. Chem. C, Vol. 112, No. 37, 2008 14643 Finally, the same methodology was applied to compare the photoresponse of ZnO nanowires and ZnO thin films.36 In general, higher photocurrent values Iph were obtained with larger grains,37-41 due to increasing mobility which approaches bulk behavior (∼200 cm2/(V s)) in high-quality thin films,42 with TF typical photoconductive gain values of Gph,lit ) 1360.38 Using the polarization and geometric conditions reported by the authors TF TF (Vph,lit ) 5 V; Lph,lit ) 10 µ), the normalized photoconductive TF gain was found to be gph,lit ) 3 × 10-8 m2/V, whose value is clearly lower than the one reported with single ZnO nanowires. On the contrary, the normalized gain per bandwidth of ZnO TF thin films is significantly higher (gph,lit BW ) 3 × 10-3 m2 Hz/ NW V) than that obtained with ZnO nanowires (gph,lit BW ) 5 × TF -8 2 10 38 10 m Hz/V ), if τlit ≈ 1.5 µs. Comparing this result with eqs 14 and 15, it is concluded that the higher photogain achieved with individual nanowires is mainly associated to the longer lifetime of the photocarriers, which increases at the expense of diminishing dynamic response. Conclusions We presented in detail the principles ruling the response of UV photodetectors based on metal oxide nanowires. Different design and fabrication strategies to enhance their performances were identified and discussed, such as controlling their geometry or tuning the intrinsic electrical properties of nanowires. Most of them were validated with experimental results obtained with photodetectors based on single ZnO nanowires. Finally, a rigorous methodology to compare different devices was presented, overcoming the present lack of systematic study in this field. On the basis of this methodology we conclude that current photodetectors based on single ZnO nanowires achieve better normalized gains (up to 3 orders of magnitude) with slower response compared with thin film devices. Acknowledgment. This work has been partially supported by the Spanish Ministry of Education (MEC) through the project N-MOSEN (MAT2007-66741-C02-01), the UE through the project NAWACS (NAN2006-28568-E), the Human Potential ProgramsAccess to Research Infrastructuress, and the project MAGASENS and CROMINA. J.D.P. and R.J.D. are indebted to the MEC for the FPU grant. F.H.-R. is indebted to the MEC for the FPU grant and the support of the Torres Quevedo program (PTQ05-02-03201). Thanks are due to the German Science Foundation (DFG) for supporting this work in the frame of priority on nanomaterialssSonderforschungsbereich 277sat the Saarland University, Saarbruecken, Germany. Supporting Information Available: (I) Photogeneration theory details: nonuniform absorption and transient solution and (II) self-heating effects on nanowires. This information is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Kind, H.; Yan, H.; Messer, B.; Law, M.; Yang, P. AdV. Mater. 2002, 14, 158–160. (2) Keem, K.; Kim, H.; Kim, G. T.; Lee, J. S.; Min, B.; Cho, K.; Sung, M. Y.; Kim, S. Appl. Phys. Lett. 2004, 84, 4376–4378. (3) Fan, Z. Y.; Chang, P. C.; Lu, J. G.; Walter, E. C.; Penner, R. M.; Lin, C. H.; Lee, H. P. Appl. Phys. Lett. 2004, 85, 6128–6130. (4) Heo, Y. W.; Kang, B. S.; Tien, L. C.; Norton, D. P.; Ren, F.; La Roche, J. R.; Pearton, S. J. Appl. Phys. A: Mater. Sci. Process. 2005, 80, 497–499. (5) Kumar, S.; Gupta, V.; Sreenivas, K. Nanotechnology 2005, 16, 1167–1171. (6) Hsu, C. L.; Chang, S. J.; Lin, Y. R.; Li, P. C.; Lin, T. S.; Tsai, S. Y.; Lu, T. H.; Chen, I. C. Chem. Phys. Lett. 2005, 416, 75–78. (7) Law, J. B. K.; Thong, J. T. L. Appl. Phys. Lett. 2006, 88, 133114.

14644 J. Phys. Chem. C, Vol. 112, No. 37, 2008 (8) Luo, L.; Zhang, Y. F.; Mao, S. S.; Lin, L. W. Sens. Actuators, A 2006, 127, 201–206. (9) Suehiro, J.; Nakagawa, N.; Hidaka, S.; Ueda, M.; Imasaka, K.; Higashihata, M.; Okada, T.; Hara, M. Nanotechnology 2006, 17, 2567– 2573. (10) Soci, C.; Zhang, A.; Xiang, B.; Dayeh, S. A.; Aplin, D. P. R.; Park, J.; Bao, X. Y.; Lo, Y. H.; Wang, D. Nano Lett. 2007, 7, 1003–1009. (11) (a) Mathur, S.; Barth, S.; Shen, H.; Pyun, J. C.; Werner, U. Small 2005, 1, 713–717. (b) Mathur, S.; Barth, S. Small 2007, 3, 2070–2075. (12) Wang, Z. L. J. Phys.: Condens. Matter 2004, 16, R829-R858. (13) Herna´ndez-Ramı´rez, F.; Rodrı´guez, J.; Casals, O.; Russinyol, E.; Vila`, A; Romano-Rodrı´guez, A.; Morante, J. R.; Abid, M. Sens. Actuators, B 2006, 118, 198–203. (14) Herna´ndez-Ramı´rez, F.; Taranco´n, A.; Casals, O.; Rodrı´guez, J.; Romano-Rodrı´guez, A.; Morante, J. R.; Barth, S.; Mathur, S.; Choi, T. Y.; Poulikakos, D.; Callegari, V.; Nellen, P. M. Nanotechnology 2006, 17, 5577– 5583. (15) Hernandez-Ramirez, F.; Tarancon, A.; Casals, O.; Pellicer, E.; Rodriguez, J.; Romano-Rodriguez, A.; Morante, J. R.; Barth, S.; Mathur, S. Phys. ReV. B 2007, 76, 085429. (16) Hernandez-Ramirez, F.; Prades, J. D.; Tarancon, A.; Barth, S.; Casals, O.; Jime´nez-Diaz, R.; Pellicer, E.; Rodriguez, J.; Juli, M. A.; Romano-Rodriguez, A.; Morante, J. R.; Mathur, S.; Helwig, A.; Spannhake, J.; Mueller, G. Nanotechnology 2007, 18, 495501. (17) Zhang, Z.; Yao, K.; Liu, Y.; Jin, C.; Liang, X.; Chen, Q.; Peng, L. M. AdV. Funct. Mater. 2007, 17, 2478–2489. (18) (a) Rose, A. Concepts in PhotoconductiVity and Allied Problems; Interscience Publishers: New York, 1963. (b) Sze, S. M. Physics of Semiconductor DeVices; John Wiley & Sons, Inc: New York, 1981. (c) Bube, R. H. Photoelectronic Properties of Semiconductors; Cambridge University Press: Cambridge, 1992. (19) Yoshikawa, H.; Adachi, S. Jpn. J. Appl. Phys 1997, 36, 6237– 6243. (20) Wang, Z. L. Mat. Today 2004, 26–33. (21) Kumar, S.; Rajaraman, S.; Gerhardt, R. A.; Wang, Z. L.; Hesketh, P. J. Electrochim. Acta 2005, 51, 943–951. (22) Suehiro, J.; Nakagawa, N.; Hidaka, S.; Ueda, M.; Imasaka, K.; Higashihata, M.; Okada, T.; Hara, M. Nanotechnology 2006, 17, 2567– 2573.

Prades et al. (23) This resolution value was taken as a rough approximation to the minimal feature of current lithography techniques. (24) Zheng, X. G.; Li, Q. Sh.; Zhao, J. P.; Chen, D.; Zhao, B.; Yang, Y. J.; Zhang, L. Ch. Appl. Surf. Sci. 2006, 253, 2264–2267. (25) Jie, J. S.; Zhang, W. J.; Jiang, Y.; Meng, X. M.; Li, Y. Q.; Lee, S. T. Nano Lett. 2006, 6, 1887–1892. (26) Studenikin, S. A.; Golego, N.; Cocivera, M. J. Appl. Phys. 2000, 87, 2413–2421. (27) Lao, C. S.; Park, M. C.; Kuang, Q.; Deng, Y.; Sood, A. K.; Polla, D. L.; Wang, Z. L. J. Am. Chem. Soc. 2007, 129, 12096–12097. (28) Chang, P. C.; Chien, C. J.; Stichtenoth, D.; Ronning, C.; Lu, J. G. Appl. Phys. Lett. 2007, 90, 113101. (29) Fan, Z. Y.; Wang, D. W.; Chang, P. C.; Tseng, W. Y.; Lu, J. G. Appl. Phys. Lett. 2004, 85, 5923. (30) Chang, P. C.; Fan, Z.; Wang, D.; Tseng, W. Y.; Chiou, W. A.; Hong, J.; Lu, J. G. Chem. Mater. 2004, 16, 5133–5137. (31) Chang, P.; Fan, Z.; Chien, C.; Stichtenoth, D.; Ronning, C.; Lu, J. G. Appl. Phys. Lett. 2006, 89, 133113. (32) Park, W. I.; Kim, J. S.; Yi, G.-C.; Bae, M. H.; Lee, H.-J. Appl. Phys. Lett. 2004, 85, 5052–5054. (33) Hong, W.-K.; Kimm, B.-J.; Kim, T.-W.; Jo, G.; Song, S.; Kwon, S.-S.; Yoon, A.; Stach, E. A.; Lee, T. Colloids Surf. A 2008, 313-314, 378–382. (34) He, J. H.; Lin, Y. H.; McConney, M. E.; Tsuruk, V. V.; Wang, Z. L.; Bao, G. J. Appl. Phys. 2007, 102, 084303. (35) Hu, L.; Chen, G. Nano Lett. 2007, 7, 3249–3252. (36) Monroy, E.; Omnes, F.; Calle, F. Semicond. Sci. Technol. 2003, 18, R33-R51. (37) Takahashi, Y.; Kanamori, M.; Kondoh, A.; Minoura, H.; Ohya, Y Jpn. J. Appl. Phys 1994, 33, 6611–6615. (38) Liu, Y.; Gorla, C. R.; Liang, S.; Emanetoglu, N.; Lu, Y.; Shen, H.; Wraback, M. J. Electron. Mater. 2000, 29. (39) Basak, D.; Amin, G.; Mallik, B.; Paul, G. K.; Sen, S. K. J. Cryst. Growth 2003, 256, 73–77. (40) Liu, M.; Kim, H. K. Appl. Phys. Lett. 2004, 84, 173–175. (41) Liu, J.; Xia, Y.; Wang, L.; Su, Q.; Shi, W. Appl. Surf. Sci. 2006, 253, 5218–5222. (42) Ohtomo, A.; Tsukazaki, A. Semicond. Sci. Technol. 2005, 20, S1-S12.

JP804614Q